On the Performance of Differential Evolution Variants in Constrained Structural Optimization
| Autor: | Vagelis Plevris, Manolis Georgioudakis |
|---|---|
| Rok vydání: | 2020 |
| Předmět: |
Sade
0209 industrial biotechnology Mathematical optimization Optimization problem Computer science Jde Population 02 engineering and technology Structural optimization Industrial and Manufacturing Engineering 020901 industrial engineering & automation 0203 mechanical engineering Jade Artificial Intelligence Robustness (computer science) Search algorithm education education.field_of_study Code Constrained optimization De Nonlinear system 020303 mechanical engineering & transports Differential evolution Benchmark (computing) |
| Zdroj: | Procedia Manufacturing |
| ISSN: | 2351-9789 |
| Popis: | Constrained optimization is a highly important field of engineering as most real-world optimization problems are associated with one or several constraints. Such problems are often challenging to solve due to their complexity and high nonlinearity. Differential evolution (DE) is arguably one of the most versatile and stable population-based search algorithms that exhibits robustness to multi-modal problems and has shown to be very efficient when solving constrained global optimization problems. In this paper we investigate the performance of several DE variants existing in the literature such as the traditional DE, the composite DE (CoDE), the adaptive DE with optional external archive (JADE) and the self-adaptive DE (jDE and SaDE), for handling constrained structural optimization problems. The performance of each DE variant is quantified by using three well-known benchmark structures in 2D and 3D. It is shown that JADE, which updates control parameters in an adaptive way, truly exhibits superior performance and outperforms the other DE variants in all the cases examined. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|